Job Scheduling Based on Dependent and Independent Task using Particle Swarm Optimization
Rajinder kaur1, Kiranbir kaur2

1Rajinder kaur*, Department of Computer Science & Technology, Guru Nanak Dev University, Amritsar, India.
2Kiranbir kaur, Assistant Professor, Department of Computer Science & Technology, Guru Nanak Dev University, Amritsar, India.
Manuscript received on March 15, 2020. | Revised Manuscript received on March 27, 2020. | Manuscript published on April 10, 2020. | PP: 1600-1606 | Volume-9 Issue-6, April 2020. | Retrieval Number: F4511049620/2020©BEIESP | DOI: 10.35940/ijitee.F4511.049620
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Mobile Cloud Computing is an accumulation of both Cloud Computing and Mobile Computing. In cloud computing resources are deployed to a client on-demand basis. Mobile cloud computing is similar to cloud computing except that some devices involved in mobile cloud computing should be mobile. The demand for MCC has been increasing due to its scalability, reliability, high QOS (Quality Of Services), longer battery life, large storage capacity. Mobile cloud computing aims to take benefit of limited resources provided by a cloud provider. Task scheduling is a major concept involved in executing a task. In cloud computing job scheduling is required to execute each job without any deadlock. But the scheduling of dependent tasks is a problem in cloud systems. This problem is an NP-complete problem and can be solved using various heuristic and metaheuristic approaches. These approaches give high-quality solutions with reasonable execution time. Particle Swarm Optimization (PSO) is one of these meta-heuristic approaches that solve the problem of grid scheduling. In this paper, we address the problem encounter in dynamic scheduling. In dynamic scheduling, each task has its own deadline completion time. The task that arrived earlier in the system occupied the resources first and later arrived tasks are rejected because their execution time exceeds the deadline. In this paper, we proposed PSO with a variable job identifier that identifies independent and dependent tasks from the population. The particles are arranged with a grid dynamically and influence swarm to minimize execution time and waiting time simultaneously. The experimental studies show that the proposed approach is more efficient than other PSO based approaches as described in the literature. 
Keywords: Fault Tolerance Rate, Particle Swarm Optimization, Virtual Machine Optimization, Cloudlets.
Scope of the Article: Machine Learning (ML) and Knowledge Mining (KM)